v0.8.0

What's Changed

• fix: LLM endpoint port update by @wewang58 in #612
• Adding Axlearn docs to adoption section in site by @Edwinhr716 in #616
• docs: Add llm-d to adoption section in site by @yankay in #617
• Bump github.com/onsi/ginkgo/v2 from 2.23.4 to 2.25.1 by @dependabot[bot] in #624
• Bump github.com/stretchr/testify from 1.10.0 to 1.11.0 by @dependabot[bot] in #625
• style: 🔨 Maintain a uniform +kubebuilder style in the context. by @yafengio in #626
• fix: 🐛 Declare the variable CGO_ENABLED in Dockerfile by @yafengio in #623
• Bump github.com/onsi/gomega from 1.38.0 to 1.38.1 by @dependabot[bot] in #627
• Bump the kubernetes group with 8 updates by @dependabot[bot] in #628
• Reverse leader and worker name by @wewang58 in #629
• KEP-622: VolumeClaimTemplates support for Leader and Worker StatefulSets by @andyzhangx in #636
• test: add trivy github action to scan image CVE in every PR by @andyzhangx in #643
• Bump github.com/onsi/ginkgo/v2 from 2.25.1 to 2.25.3 by @dependabot[bot] in #639
• test: fix trivy action by @andyzhangx in #646
• Update k8s dependencies to 1.34.0 by @ardaguclu in #647
• chore: add buildDate in driver init logs by @andyzhangx in #642
• test: set up unit test coverage by @andyzhangx in #648
• Validate partition field be non-negative by @LuyuZhang00 in #618
• Bump github.com/open-policy-agent/cert-controller from 0.13.0 to 0.14.0 by @dependabot[bot] in #637
• Bump github.com/stretchr/testify from 1.11.0 to 1.11.1 by @dependabot[bot] in #640
• chore: add version parameter by @andyzhangx in #650
• feat: add volumeClaimTemplates and PersistentVolumeClaimRetentionPolicy in LeaderWorkerSetSpec by @andyzhangx in #633
• Bump the kubernetes group with 8 updates by @dependabot[bot] in #651
• Run make generate and make manifests to update CRD by @ardaguclu in #655
• Add Red Hat in adopters by @ardaguclu in #656
• doc: add volumeClaimTemplates example by @andyzhangx in #649
• docs(concepts): add Startup Policy section and example by @Nihat-AYDIN in #661
• test: add codespell github action by @andyzhangx in #660
• docs: expand SubGroup section with size and LeaderOnly type by @tsj-30 in #662
• Bump versions to fix failing e2e tests by @Edwinhr716 in #674
• chore(deps): bump github.com/onsi/ginkgo/v2 from 2.25.3 to 2.26.0 by @dependabot[bot] in #667
• [bugfix] update helm version by @LuyuZhang00 in #672
• [cleanup] Update to golang 1.25 by @yankay in #678
• Use explicit nginx image path instead of ambiguous shortname by @ardaguclu in #679
• [Docs] Updating vllm versions by @Edwinhr716 in #680
• Add Kubeflow Trainer to the LWS Integrations by @andreyvelich in #681
• docs: standardize Hugging Face env vars to HF_TOKEN by @yankay in #683
• chore(deps): bump github.com/onsi/ginkgo/v2 from 2.26.0 to 2.27.1 by @dependabot[bot] in #687
• chore(deps): bump sigs.k8s.io/controller-runtime from 0.22.3 to 0.22.4 by @dependabot[bot] in #691
• chore(deps): bump github.com/onsi/ginkgo/v2 from 2.27.1 to 2.27.2 by @dependabot[bot] in #692
• Fix: drop UPDATE from pod mutating webhook by @ycjiang50 in #697
• Fix e2e test assertions and resource cleanup by @ErikJiang in #701
• feat: Enable certain features of the scheduler using annotations (#669) by @wangshulei098 in #670
• fix makefile crds target and broken links in docs by @ErikJiang in #700
• chore(deps): bump the kubernetes group with 8 updates by @dependabot[bot] in #702
• chore(deps-dev): bump autoprefixer from 10.4.21 to 10.4.22 in /site in the all group by @dependabot[bot] in #703
• chore(deps): bump github.com/open-policy-agent/cert-controller from 0.14.0 to 0.15.0 by @dependabot[bot] in #708
• chore(deps): bump sigs.k8s.io/structured-merge-diff/v6 from 6.3.0 to 6.3.1 by @dependabot[bot] in #709
• Adding lws+kueue user guide by @Edwinhr716 in #682
• Upgrade Go to 1.25 to fix Trivy CVEs by @yankay in #713
• Upgrade Hugo to v0.152.2 and migrate site to Docsy Go module by @yankay in #714
• chore(deps): bump github.com/onsi/ginkgo/v2 from 2.27.2 to 2.27.3 by @dependabot[bot] in #716
• chore(deps): bump the kubernetes group with 8 updates by @dependabot[bot] in #719
• chore(deps-dev): bump autoprefixer from 10.4.22 to 10.4.23 in /site in the all group by @dependabot[bot] in #721
• chore(deps): bump github.com/onsi/gomega from 1.38.2 to 1.38.3 by @dependabot[bot] in #720
• Injecting environment variables for new TPU machine by @Edwinhr716 in #722
• Try to fix Helm so it does not allow deleting CRDs by @Mostafahassen1 in #718
• Add RecreateGroupAfterStart annotation by @Edwinhr716 in #725
• Changing recreateGroupAfterStart annotation to be snake case instead by @Edwinhr716 in #728
• Updating docs for MaxUnavailable now that it is in Beta by @Edwinhr716 in #727
• Adding docs for RecreateGroupAfterStart by @Edwinhr716 in #729
• Returning if the leader statefulset is being deleted by @Edwinhr716 in #733
• chore(deps): bump github.com/onsi/ginkgo/v2 from 2.27.3 to 2.27.5 by @dependabot[bot] in #735
• chore(deps): bump github.com/onsi/gomega from 1.38.3 to 1.39.0 by @dependabot[bot] in #730
• fix(manifests): Fix common labels for Kustomize deployment by @andreyvelich in #739
• Enable certs manager readiness check by @j-skiba in #737
• Release 0.8.0 by @Edwinhr716 in #740

New Contributors

• @wewang58 made their first contribution in #612
• @yafengio made their first contribution in #626
• @andyzhangx made their first contribution in #636
• @LuyuZhang00 made their first contribution in #618
• @Nihat-AYDIN made their first contribution in #661
• @tsj-30 made their first contribution in #662
• @andreyvelich made their first contribution in #681
• @ycjiang50 made their first contribution in #697
• @ErikJiang made their first contribution in #701
• @wangshulei098 made their first contribution in #670
• @Mostafahassen1 made their first contribution in #718
• @j-skiba made their first contribution in #737

Full Changelog: v0.7.0...v0.8.0

v0.7.0

What's Changed

🚀 Major Features:

• add kep-511 partition update by @lixin010 in #541
• KEP #552: allow updates to spec.leaderWorkerTemplate.size field by @synthe102 in #557
• Add KEP#407: Gang Scheduling by @JesseStutler in #496

✨ Features:

• Add LWS configuration in Helm Charts by @ardaguclu in #517
• feat: add meaningful printer columns to LeaderWorkerSet CRD by @xigang in #580

🐛 Bugs:

• Fix helm update by @kerthcet in #481
• Fix UpdateInProgress conditions for newly created LWS by @yankay in #490
• Fix helm install panic by invalid values .Values.certManager.enable by @nicole-lihui in #520
• Fix issue with rolling update when some existing replicas are unhealthy by @pierewoj in #488
• Fix lws chart: the webhook service selects unexpected pods when it is introduced as a dependency by @carlory in #536
• fix: support foreground deletion by @congcongke in #523
• fix: llamacpp compile error with adding curl dependency by @yankay in #559
• fix rolling update stuck that when old pods were not ready by @gujingit in #568

♻️ Cleanups:

• use a bigger machine for cloud build by @kannon92 in #483
• [CI] Add more verify include gomod, fmt, and generate-apiref by @yankay in #485
• docs: remove the use of Downward API for LWS_WORKER_INDEX by @chewong in #486
• [Docs] Fixed location of vLLM script on TPU image by @Edwinhr716 in #494
• Introduce env variable to modify ns in metrics tests by @ardaguclu in #489
• Add an example of using OCI to the Helm installation guide by @yankay in #503
• Add prepare-release-branch in Makefile to update versions automatically by @yankay in #504
• Inject certs for metrics endpoint when cert-manager is enabled by @ardaguclu in #478
• Remove duplicate tag key in chart values by @nicolasochem in #509
• Run e2e tests also when cert-manager is enabled by @ardaguclu in #518
• Removing SaxML docs by @Edwinhr716 in #522
• Use correct lws apigroup name by @ardaguclu in #527
• fix: check optional field not nil before reference by @panpan0000 in #530
• Example of Configuration CRD by @panpan0000 in #526
• Add dynamo and new presentation by @kerthcet in #549
• test: fix race condition in worker StatefulSet deletion test case by @xigang in #574
• refactor: remove unused lws parameter from CreateRevision function by @xigang in #573
• Update vllm.md by @jonathan-innis in #581
• [adoption] add OME as new adopter by @panpan0000 in #579
• [talk and slides] adding KubeCon HK and JP 2025 talks for LWS by @panpan0000 in #582
• docs: Add HUGGING_FACE_HUB_TOKEN as environment variable by @jonathan-innis in #583
• auto-sync the CRD via make generate & make helm-chart-push by @kerthcet in #592
• Default Partition in RollingUpdateConfiguration to 0 to align with spec by @ardaguclu in #594
• fix: disable CURL to resolve libcurl dependency issues in llamacpp example by @yankay in #599

New Contributors

• @nicolasochem made their first contribution in #509
• @panpan0000 made their first contribution in #525
• @pierewoj made their first contribution in #488
• @lixin010 made their first contribution in #541
• @synthe102 made their first contribution in #557
• @xigang made their first contribution in #574
• @jonathan-innis made their first contribution in #581
• @JesseStutler made their first contribution in #496

Full Changelog: v0.6.0...v0.7.0

v0.6.3

What's Changed

• Fix: use the main image in helm chart by @kerthcet in #597

Full Changelog: v0.6.2...v0.6.3

v0.6.2

What's Changed

Fix lws chart: the webhook service selects unexpected pods when it is introduced as a dependency #536

v0.6.1

What's Changed

• Fix UpdateInProgress conditions for newly created LWS by @yankay in #490

v0.6.0

What's Changed:

🚀 Major Features:

• Introduce new component config flag by @ardaguclu in #325
• Creating LWS site by @Edwinhr716 in #426
• KEP 257: LeaderExcluded SubGroup support by @Edwinhr716 in #428

✨ Features:

• Make lws image multi-platform by @yankay in #321
• Inject environment variable LWS_WORKER_INDEX by @singeleaf in #416

🐛 Bugs:

• Add controllerrevisions rbac to helm chart. by @wqlparallel in #319
• Fix for odd numbers of arguments passed as key-value pairs for logging issue by @KeerthanaAP in #360
• Fix infinite StatefulSet creation loops by validate leader annotations by @yankay in #394
• Change annotation leaderworkerset.gke.io/subgroup-size to leaderworkerset.sigs.k8s.io/subgroup-size by @yankay in #434
• ignore 0 replica check by @congcongke in #439

♻️ Cleanups:

• Remove pods create permission by @wqlparallel in #323
• add junit and artifacts directory for e2e tests so testgrid will disp… by @kannon92 in #335
• Pass ARTIFACTS env variable when building e2e tests by @Edwinhr716 in #345
• Add adopters, integrations, talks by @kerthcet in #367
• Deprecate flags as the recommended path is to pass configuration file by @ardaguclu in #361
• doc: Add terraform based EKS Blueprints pattern for multi-node inference by @askulkarni2 in #373
• Add missing external types to apply configurations by @astefanutti in #374
• Ignore LeaderElectionNamespace in cmd unit tests as it is dynamic by @ardaguclu in #383
• Allow pass through of additional params to "ray start" in ray_init.sh by @dims in #378
• use staging images instead of gcr by @kannon92 in #398
• Add files to publish helm chart by @Edwinhr716 in #411
• fix image build toolchain errors by @kannon92 in #424
• Run controller-gen only for api and pkg directories by @ardaguclu in #417
• Add Label leaderworkerset.sigs.k8s.io/name to the headlessService by @yankay in #433
• Use create jobs command instead of run pods in e2e testing by @ardaguclu in #445
• Export kind and cluster related logs under artifacts directory for e2e by @ardaguclu in #451
• Set readOnlyRootFilesystem explicitly to true by @ardaguclu in #457
• Check controller_runtime_webhook_requests_total that always exists in both pods by @ardaguclu in #456
• Explicitly bind metrics 8443 as containerPort in kustomize and helm by @ardaguclu in #458

📃 Docs:

• Adding User Guide for TensorRT-LLM by @Edwinhr716 in #356
• Adding User Guide for SGLang by @yankay in #377
• [Docs] Use the vllm image instead of building by the user self in the sample by @yankay in #453
• [Doc] Add troubleshooting page in site by @yankay in #468

New Contributors

• @yankay made their first contribution in #321
• @ardaguclu made their first contribution in #325
• @DiegoLelis made their first contribution in #354
• @KeerthanaAP made their first contribution in #360
• @askulkarni2 made their first contribution in #373
• @astefanutti made their first contribution in #374
• @dims made their first contribution in #378
• @singeleaf made their first contribution in #416
• @congcongke made their first contribution in #439

Full Changelog: v0.5.0...v0.6.0

v0.5.1

What's Changed

• Automated cherry pick of #347: Removed update operation from pod webhook by @Edwinhr716 in #348
• Automated cherry pick of #342: Added handling of nil revision in pod controller by @Edwinhr716 in #349
• Automated cherry pick of #341: Removed restarting group log in pod controller by @Edwinhr716 in #350

Full Changelog: v0.5.0...v0.5.1

Release v0.5.0

What's Changed

• Update installation guide by @liurupeng in #217
• Bump the kubernetes group with 5 updates by @dependabot in #224
• Add support to install lws at an arbitrary namespace by @kerthcet in #226
• feat: add build arg to docker buildx by @vie-serendipity in #227
• Injecting TPU_NAME environment variable by @Edwinhr716 in #209
• Creating new workerStatefulSet instead of patching when updating LWS template by @Edwinhr716 in #229
• Update KEP #173 to reflect implementation by @Edwinhr716 in #233
• Add cherry-pick script by @liurupeng in #235
• Add vLLM TPU Multihost example by @Edwinhr716 in #234
• Bump the kubernetes group with 5 updates by @dependabot in #241
• Bump sigs.k8s.io/controller-runtime from 0.19.0 to 0.19.1 by @dependabot in #242
• Changed log level to 2 by @Edwinhr716 in #244
• Bump github.com/onsi/ginkgo/v2 from 2.20.2 to 2.21.0 by @dependabot in #245
• Bump github.com/onsi/gomega from 1.34.2 to 1.35.1 by @dependabot in #246
• change to not use NetworkConfig in hash calculation if it has default value by @Edwinhr716 in #250
• Added validation for subdomainPolicy values by @Edwinhr716 in #251
• changed cherry pick script to be executable by @Edwinhr716 in #253
• add pods finalizer update rbac to allow lws to run in openshift by @kannon92 in #249
• Update installation instructions for release 0.4.2 by @ahg-g in #255
• Bump github.com/open-policy-agent/cert-controller from 0.11.0 to 0.12.0 by @dependabot in #256
• Bump sigs.k8s.io/structured-merge-diff/v4 from 4.4.1 to 4.4.3 by @dependabot in #258
• Unify Leader Election Namespace with Deployment Namespace by @wqlparallel in #263
• Bump github.com/onsi/gomega from 1.35.1 to 1.36.0 by @dependabot in #271
• Added validation for LWS Name by @Edwinhr716 in #272
• Updating OWNERS file by @Edwinhr716 in #273
• Bump github.com/onsi/ginkgo/v2 from 2.21.0 to 2.22.0 by @dependabot in #269
• Bump sigs.k8s.io/controller-runtime from 0.19.1 to 0.19.2 by @dependabot in #270
• Bump the kubernetes group with 5 updates by @dependabot in #268
• Bump sigs.k8s.io/controller-runtime from 0.19.2 to 0.19.3 by @dependabot in #278
• update golang to 1.23 and fix docker warnings by @kannon92 in #279
• Use the namespace of the service account by @kannon92 in #282
• Bump the kubernetes group with 5 updates by @dependabot in #286
• Bump sigs.k8s.io/structured-merge-diff/v4 from 4.4.3 to 4.5.0 by @dependabot in #287
• Bump github.com/onsi/gomega from 1.36.0 to 1.36.1 by @dependabot in #288
• add doc to say how to install in a different namespace by @kannon92 in #285
• Add helm charts for lws. by @wqlparallel in #266
• add KEP 238, to add controller revision by @Edwinhr716 in #261
• address non linear parsing cve by upgrading golang net by @kannon92 in #290
• Bump github.com/onsi/ginkgo/v2 from 2.22.0 to 2.22.1 by @dependabot in #292
• Remove namespace arg from manager command by @chiragjn in #293
• Add Controller Revision (Implementation of KEP #238) by @Edwinhr716 in #277
• Bump github.com/onsi/gomega from 1.36.1 to 1.36.2 by @dependabot in #295
• Tests for Controller Revision by @Edwinhr716 in #297
• Minimize number of unnecessary logs by @Edwinhr716 in #298
• Update Reference Doc by @Edwinhr716 in #300
• Updating owners by @Edwinhr716 in #301
• Add Events by @Edwinhr716 in #299
• Bump github.com/onsi/ginkgo/v2 from 2.22.1 to 2.22.2 by @dependabot in #303
• Replaced Kueue with LWS on chart/lws ReadMe by @Edwinhr716 in #305
• Ignore not found error when handling ReacreateGroupOnPodRestart by @Edwinhr716 in #304
• Replace kube-rbac-proxy by @anshuman-agarwala in #284
• Fixed flaky ControllerRevision test and add CR events by @Edwinhr716 in #306
• Fixed metrics service port, added log level flag back by @Edwinhr716 in #307
• Removed kube-rbac-proxy container from helm charts by @Edwinhr716 in #308

New Contributors

• @vie-serendipity made their first contribution in #227
• @wqlparallel made their first contribution in #263
• @chiragjn made their first contribution in #293
• @anshuman-agarwala made their first contribution in #284

Full Changelog: v0.4.0...v0.5.0

0.4.2

What's Changed

• Automated cherry pick of #250: change to not use NetworkConfig in hash calculation if it has default value by @Edwinhr716 in #252

Full Changelog: v0.4.1...v0.4.2

LeaderWorkerSet Patch release 0.4.1

This is a patch release for v0.4.0